Architecture details
The Content Localization on AWS solution consists of a number of components that are deployed to AWS using CloudFormation nested stacks. The components include the Media Insights on AWS solution, a Media Insights on AWS workflow, a web application, web application authentication components, and an Amazon OpenSearch Service MI custom pipeline consumer.
AWS services in this solution
AWS service | Description |
---|---|
Amazon API Gateway |
Core. Entry point to interact with the control and data plane APIs where requests to create, read, update, delete, or run workflows begin, or data retrieval begin. |
Amazon DynamoDB |
Core. Stores workflow-related data, such as state machine definitions for operators, workflow configurations, and workflow run statuses. |
AWS Identity and Access Management (IAM) |
Core. Grants the solution’s AWS Lambda function access to create Regional resources. |
AWS Lambda |
Core. Supports workflow orchestration, operators executions, and store workflow results. |
Amazon Simple Notification Service |
Core. Supports workflow execution rate limits. |
Amazon Simple Queue Service |
Core. Supports workflow execution rate limits. |
Amazon Simple Storage Service |
Core. Resource used for storing input user media and output transformed media by the workflow. |
AWS Step Functions |
Core. Creates a resource when a user defines a new workflow using the workflow API. |
Amazon Comprehend |
Supporting. Can be integrated into workflows to find key phrases in text and references to real-world objects, dates, and quantities in text. |
Amazon
Elemental MediaConvert |
Supporting. Can be integrated into workflows to transcode input video into MPEG-4 format and generate thumbnails. |
Amazon Kinesis Data Streams |
Supporting. Used to stream data changes reflected in DynamoDB which consumers can ETL data from the data store to downstream services. |
Amazon Polly |
Supporting. Can be integrated into workflows to turn input text into speech. |
Amazon Rekognition |
Supporting. Can be integrated into workflows for Celebrity Recognition, Content Moderation, Face Detection, Face Search, Label Detection, Person Tracking, Shot, Text, and Technical Cue Detection. |
AWS Systems Manager |
Supporting. Provides application-level resource monitoring and visualization of resource operations and cost data. |
Amazon Translate |
Supporting. Can be integrated into workflows to translate input text. |
Amazon Transcribe |
Supporting. Can be integrated into workflows to create SRT or VTT caption files from video transcripts. It can also convert input audio to text. |
AWS X-Ray |
Supporting. Provides debugging tools for the Media Insights on AWS application. |
Media Insights on AWS solution
Media Insights on AWS is deployed in one of two modes, depending on which option you choose:
-
As a nested stack from the
deployment/content-localization-on-aws.yaml
template. -
As a standalone stack whose name is provided as a parameter input to the
deployment/content-localization-on-aws-use-existing-mie-stack.yaml
template.
Media Insights on AWS provides services for creating and running the content localization workflow, and for storage and retrieval of the media objects and metadata that are generated by the workflow for each input video (asset). At deployment time, this solution integrates with Media Insights on AWS to create a workflow using the MI operator library and MI workflow custom resources. At runtime, the Content Localization on AWS solution integrates with Media Insights on AWS through its REST APIs to run the content localization workflow, and store and retrieve media objects and analysis outputs that are created by the workflow. Finally, the solution integrates with Media Insights on AWS through DynamoDB stream events emitted by the MI data pipeline as workflows run to store analysis results in Amazon OpenSearch Service.